Attitude as a concept has played a central role in social psychology for decades. Earlier definitions of what an attitude is still hold weighting in contemporary society. For example, Thurstone (1931, p. 261) defined an attitude as “the affect for or against a psychological object.” Decades later, highly influential research defined attitudes similarly as “favourable or unfavourable dispositions toward social objects, such as people, places, and policies” (Greenwald & Banaji, 1995, p. 7). Attitude formation has been a long standing issue in attitude research. There are various psychological factors involved in attitude formation. Both cognition and affect are two central components of attitudes. The cognitive basis for an attitude has been defined as beliefs, judgements and thoughts whereas the affective basis for an attitude includes emotions and feelings (McGuire, 1969). McGuire (1969) suggested that these two components affect information processing in different ways. These two components make up part of the classic tripartite model of attitude with “behaviour” being the third component (Katz & Stotland, 1959). Past research also contended that all three components must be present for an attitude to exist; a fact that has now been described as dubious (Hogg & Cooper, 2000). Contemporary research examines how one or two components can exist without the other. Cognitive and affective components of attitude will be focused on for the purpose of this paper. The question under contention is whether affect or cognition influences attitude formation more.
One vital way of forming an attitude is cognitive information about the target, i.e. beliefs about the attributes of the target. Knowledge about an attitude object may stem for direct experience with the object or from indirect sources such as parents, friends and the media. A second vital way of forming an attitude is based on how the target makes them feel and what emotions may be evoked (Weiner et al., 2003). Maio and Olson (2000) differentiate from these two components succinctly by describing affective based attitudes as feelings and cognitive based attitudes as attributes. Individuals differ in their tendency to have attitudes that are either cognitive or affect based. This can be seen in a number of self-report measures that describe individual differences e.g. ‘thinkers’ versus ‘feelers’ – those that tend to rely on their feelings or those that stick to the facts (Myers-Briggs Type Indicator, 1979). Some theoretical perspectives have focused in on the importance of affect while others emphasise the significance of cognition in forming attitudes. While it is beyond the scope of this paper to provide a comprehensive review of the literature in all of its complexity, it does aim for a better understanding of the differences between cognition and affect on attitude formation.
Empirical support for the influence of cognition on attitude formation
One of the most significant models which emphasises the role of cognition in attitude formation is Fishbein and Ajzen’s (1975) theory of reasoned action which is based on an expectancy-value model. This model distinguished conceptually the three components of attitude as mentioned earlier. However, Fishbein and Ajzen (1975) refer to these as beliefs, attitudes, and intentions with attitudes in this context implying the affective evaluation. In their opinion, cognitive structure based on an individual’s beliefs and the use of an expectancy-value model establishes a person’s attitude. There are three basic assumptions underlying this model; behavioural intentions are the only direct determinant of behaviour, behavioural intentions are determined by affective attitudes and subjective norms and affective attitudes are a function of beliefs about consequences. This model of attitude formation conceives humans as “rational, deliberate thinkers” who base their attitudes and behaviour on information they receive about a particular subject (Weiner, Freedheim, Millon & Lerner, 2003 p.311. However, this model is not without limitations as it fails to provide an explanation about the process from which an attitude comes from (Hogg & Cooper, 2003).
Attitudes are also formed in cognition using the cognitive-response model. This refers to thoughts or reactions we have in response to a message. This cognitive response approach has led to various models of persuasion. One such model is the elaboration likelihood model (Petty & Cacioppo, 1986). This model postulates that people’s motivation to cognate varies across individuals and is context dependent. It consists of two routes to persuasion; the central and the peripheral. The central route contains personal relevance and understanding. In contrast, the peripheral route contains source focus and confusion. Elaboration likelihood is said to be high “when conditions foster people’s motivation and ability to engage in issue-relevant thinking” (Petty & Cacioppo, 1986, p. 128). Cacioppo, Petty and Morris (1983) examined the effect of need for cognition on message evaluation, recall and persuasion. In this study, participants holding almost the same attitude toward instituting senior comprehensive exams who varied from high to low on the need for cognition scale read either a strong or weak argument pertaining to whether or not senior comprehensive exams should be instituted. In both their experiments they found that the strength of the argument had a greater impact on the message evaluations and source impressions provided by individuals high in need for cognition. A more recent application of the elaboration likelihood model involved expanding and testing it to encourage people to volunteer with the Special Olympics (Park, Turner & Pastor, 2008). It broadened the elaboration likelihood model by adding an emphatic tendency variable to it. Park and colleagues (2008) found that both high and low empathy participants were motivated to process the advertisements but that involvement was found to influence the argument processing of low empathy participants. This model offers person, situation and message variables that affect attitude formation (Hogg & Cooper, 2003).
Fishbein and Middlestadt (1995) documented how various theorists had begun to reject the notion of a purely cognitive basis for an attitude in the 1980’s. The authors argue that “cognition-free” processes of attitude formation should be viewed “as methodological artefacts resulting from the use of inappropriate predictors and/or criteria” (Fishbein & Middlestadt, 1995, p. 184). However, the authors did not recommend overlooking “cognition-free” processes altogether suggesting that evaluations can and do form in the absence of beliefs about the object.
Empirical support for the influence of affect on attitude formation
Attitudes that are formed from affect consist of the importance of emotional feelings attributed to an attitude object (Edwards, 1990). There exists evidence for the existence of attitude formation on affect alone. Zajonc (1980) argued that affect can be independent to and even precede cognition. Classical conditioning and mere exposure are two primary ways in which attitudes may be formed on the basis of affect (Hogg & Cooper, 2003).
Classical conditioning is a basic form of learning that was originally studied by Pavlov (1927). It postulates that when two stimuli are repeatedly associated, a person learns to respond to them with a similar emotional reaction. Classical conditioning is especially involved with the affective component of attitudes because pleasant or non-pleasant experiences can lead to pleasant or non-pleasant attitudes. For example, Olzon and Fazio (2001) provided evidence that attitudes can form implicitly through classical conditioning. Participants were told to look out for target images amongst a number of ‘random’ images on a computer screen. Amongst these random images were Pokemon cartoon characters with either positive or negative images. Tests of explicit memory and an evaluation task indicated that although the participants were not aware of the pairings, the positive images were found to be more pleasant than the negative images. This study is evidence of attitudes forming in the absence of a conscious belief about the attitude object (Olzon & Fazio, 2001).
One of the most influential contributors in this field first coined the term “mere exposure effect” (Zajonc, 1968). Mere exposure is a second way through which attitudes may be formed on the basis of affect. Infamous research carried out by Monahan, Murphy and Zajonc in 2000 had two groups of participants examine Chinese ideographs. The first group was shown the same 5 ideographs five times and the second group was shown 25 different ideographs. Results provide support for the mere exposure effect as participants who were in the repeated exposure condition indicated better moods and felt more positive.
Bodur and colleagues (2000) contributed to research emphasising the impact of affect on attitude formation. Results from their second study into the detection and prevention of HIV and AIDS and also the prevention of drink driving indicate that affect plays it own separate role in forming attitudes.
Zajonc (1980) argued for the validity of affective-based attitudes because we often doubt everything but our own feelings. How we feel rather than what objective data tells us is sometimes the most important thing. Zajonc (1980) also postulated that affective-based attitudes have implications for the self. This implies that counter attitudinal information we hear may appear as a threat to the self thus making us more motivated to defend it. Schwarz and Clore (1983) demonstrated how affect influences subsequent evaluations and attitudes when affective states are used as sources of information. According to this “How do I feel about it?” heuristic, mood is interpreted as a response to a target (Schwarz & Clore, 1983). A negative affective state that an individual experiences when thinking about an attitude object provides the individual with evidence that they have a negative attitude toward the object and vice versa, for example, good versus bad weather and whether the reason for an individual’s mood is based on the weather or not (Schwarz & Clore, 1983).
Despite these varying perspectives on what influences attitude formation more, the distinction has been described as not dichotomous (Edwards, 1990). The impact of cognitive or affective viewpoints on attitude formation depends on the nature of an attitude’s origin (Edwards, 1990). This research carried out by Edwards (1990) which primed cognitive or affective attitudes towards a beverage portrays how affect-based attitudes are more susceptible to affective means of persuasion than to cognitive means of persuasion. Moreover, cognition-based attitudes displayed equal change under both forms of persuasion. Evidence from the second experiment revealed that the expression of an attitude will be articulated with greater conviction when the attitude is affect-based. A more recent study has shown how priming either a cognitive or affective state can lead to the formation of different attitudes. Van den Berg and colleagues (2006) induced a cognitive or affective state by having participants complete a word search puzzle that contained either affectively or cognitively based words. Participants then read information about a novel animal that either contained positive affective information and negative cognitive information, or negative affective information and positive affective information. Their results indicated that when individuals were in an affective state, their evaluations of the animal were more consistent with the valence of the affective information than the cognitive information. Similarly, when individuals were in a cognitive state, their evaluations were more consistent with the valence of the cognitive information than the affective information. These findings also extended the research by looking at the possibility of activating an affective or cognitive based attitude unobtrusively.
Both cognition and affect are important contributors in the formation of attitudes. It is difficult to say which influences attitude formation more. Rather, it seems to be the case from research evidence that attitudes are formed from both cognition and affect. Eagly, Mladinic & Otto (1994) described use of the term synergistic relation to refer to the influence of affect on cognition and cognition on affect. While research has helped us to understand the role cognition and affect play, further examination of these constructs are needed.